Building upon successful previous work (N=104 heroin-dependent cases), the goal of this 5-year project is to quantify the effect on the liability to opiate dependence of specific candidate genes (including opioid, cannabinoid, dopamine and serotonin receptor genes), including gene-gene and gene-environment interactions, in a targeted sample of 2,500 opiate-dependent individuals and 2,500 controls matched on gender, ethnicity and place of birth. The CIDI, an internationally-used instrument with established reliability and validity, will be used to make DSM-IV diagnoses of opiate dependence, other substance dependence (e.g., nicotine, alcohol, benzodiazepines) and mental disorders (e.g., childhood conduct disorder, ASPD, major depressive disorder, suicidality). Under a cost-efficient strategy, the matched control sample will be generated using the anonymous DNA bank of the Laboratory of Molecular Pathology in Sofia (>80 percent of all live births in Bulgaria since 1998, >250,000 samples). Opiate abuse and dependence is a major public health concern in the United States, with an estimated 3.08 million individuals with a lifetime history of heroin use. The same is true in Bulgaria, where the proposed study would be conducted, with an estimated 30 to 50,000 heroin dependent individuals, up from approximately 1,500 in the late 1980s. Previous studies, both human and animal, have consistently demonstrated the importance of genetic factors in the development of opiate dependence. Neurobiological models for the positive and negative reinforcement effects of heroin have now been proposed, from which specific candidate genes can be identified. However, previous candidate gene studies of opiate dependence in humans have yielded conflicting results, in part due to the reliance on relatively small samples. This is especially salient when the candidate gene modulates the risk to progress from a state of controlled experimentation with heroin to a state of dependence and uncontrolled use (high transition probability, large sample size requirements). The large sample size for this study would make such genes amenable to analysis.